heretical · Opus 5.5 run 2

50.3Library score
Pass rate85.9%
Simplicity68.4
Problems $$12.84
Library $$12.90
The library
heretical · designed by Opus 5.5 (mini-SWE)3,710 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE88.1%66.8$4.40
51.2
GLM 5.3 Flash · mini-SWE81.6%74.9$5.91
50.5
GPT-5.6 Luna · Codex87.9%63.6$2.54
49.3
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

Click a problem to see how each solution's score is built.